Microsoft disrupts EvilTokens, AI scam platform that hijacked 12,000 accounts
Ars Technica AI · rss · 2026-09-23
Microsoft said it led an industry-wide disruption of EvilTokens, a subscription-based scam platform that compromised 12,000 Microsoft accounts in a few months. Launched via Telegram in February at $1,500 upfront plus $500/month, its core was an AI chatbot that analyzed victims' inboxes, identified trusted relationships and payment authorizations, prioritized high-payout targets, recommended fraud strategies, and drafted impersonation emails — compressing attacks from days to minutes.
More from Safety
- AI-Powered Flock Cameras in 9 Colorado High School Lots Spark Parental Outcry — Polymarket · 2026-09-23
- OpenAI reveals model wrote its own jailbreak instructions into compaction summaries — conitzer · 2026-09-23
- Can AI Be Slowed Down? Stanford HAI Experts Debate Agent Emergence, Self-Improvement and Kill Switches — StanfordHAI · 2026-09-23
- Claude Opus 5.5 system card: model took likely-harmful actions in ~half of security exercise runs — rohanpaul_ai · 2026-09-23
- Anthropic: models that can automate AI research need a higher safety bar — ChrisGPT · 2026-09-23
- New polling: US voters across parties reject the 'AI safety is a hoax' claim 2-to-1 — DavidSKrueger · 2026-09-23